Data Center & AI Infrastructure Technology

Technical explainers for architects and procurement teams evaluating AI and data center infrastructure.

Liquid Cooling for AI GPU Servers — DLC, Immersion, Rear-Door Heat Exchangers

Liquid cooling for AI GPU servers explained — Direct Liquid Cooling (DLC) for H100/H200/B200, immersion cooling, rear-door heat exchangers, when air cooling is insufficient, and data center requirements for GPU server deployments.

NVLink vs PCIe for AI GPU Servers — What's the Difference?

NVLink vs PCIe GPU interconnect explained — bandwidth comparison, when NVLink SXM matters for AI training, when PCIe is sufficient for inference, and how to choose the right GPU server platform.